A Novel Face Recognition Algorithm Based on Numerical value of Facial Feature Points

To improve the speed of querying the relevant information of the face by inputting the face image puts forward a face recognition method of numerical value based on the face facial feature points distance ratio. First, the face image is collected through the camera equipment and preprocessing image. Second, perform face detection and key point location on the image, and select some main key points. Third, calculate the distance between every two points, choose n distances from it that have more influence on the face difference, calculate the ratio of this n distance, and use it as a data feature of the human face. Comparison with the search time of the deep convolutional neural network of the face images set. It can see that extracting the seven feature points of the facial features and converting the face image into twenty-one-dimensional data can effectively promote the human face retrieval speed while ensuring accuracy.